Board index FlightGear The FlightGear project

Flight Dynamics Model

Questions about the FlightGear organisation, website, wiki etc.

Flight Dynamics Model

Postby awtudor » Fri Jul 06, 2012 11:31 pm

Hi,

I am currently doing research relating to neural networks and their ability to control physical systems. I am trying to set up a simulation using Flight Gear/Matlab Simulink/NI LabVIEW/neural network all put into a feedback loop (http://i.imgur.com/7ukvP.png). So far I have a couple questions.

1) I am trying to find a very realistic aircraft flight dynamics model. Something that is as best of an approximation to the real system as possible. Is there a place that details the accuracy of models that I can use (more so than 3/5, 4/5, and so on http://www.flightgear.org/download/aircraft-v2-6/)? Ideally, I would like more information on how the models were built, what kind of data was used to make them, where the flight dynamics data came from and so on. Right now I have found the Spitfire to be the best option. I am also interested in UAV models - does that exist anywhere?

2) I posted another question at viewtopic.php?f=36&t=16835 relating to the interface between Matlab Simulink and Flight Gear. I know this was posted yesterday but since this is a forum, is there any guarantee that my question will get answered? I'll also send the question to Mathworks.

3) Is there a specific person I can contact directly that could be better suited to answer my questions or is here the best place? Feel free to email me.

Thanks,
Andy

Andrew W. Tudor
awtudor2@ucla.edu
University of California, Los Angeles
Graduate Student in Mechanical and Aerospace Engineering
University of Illinois, Urbana-Champaign
B.S. Aerospace Engineering
awtudor
 
Posts: 5
Joined: Mon Jun 25, 2012 7:48 pm

Re: Flight Dynamics Model

Postby Sealbhach » Sat Jul 07, 2012 3:49 am

Hi, you'll probably get a better response if you ask on the developer mailing list: http://wiki.flightgear.org/Mailing_list

.
Sealbhach
 
Posts: 934
Joined: Wed Jun 30, 2010 10:17 am

Re: Flight Dynamics Model

Postby Gijs » Sat Jul 07, 2012 11:20 am

Hi Andrew,

sounds like an interesting project! I've tried to answer your questions:

1) The detailed models usually have some references inside the FDM files (in Aircraft/p51d/p51d-jsbsim.xml for example, you'll find some comments like "from NACA P-51X and P-51B test data" or "The rest of the drag functions are guesses."). JSBSim supports some "reference/documentation tags", but it isn't used very often, in the FlightGear aircraft at least. But the comments as the P51D has them work basically the same. The P51D has one of the better FDMs btw.

UAVs are rare as of yet. You can download the "ATI-Resolution" from the download page. I'm working on a Gatewing X100 myself, but the FDM is not scientific yet (hopefully we'll get some windtunnel data next year).

What kind of aircraft do you need for your project? Or can it be anything that flies?

2) We don't give reply-guarentees, but "normal" usage topics get a reply within hours. More "non-average-user" questions, like interfacing and source stuff takes longer, as there are only a few core developers lurking around on the forum. Like Sealbhach said, the devel-mailinglist is usually the best place for such questions.

It does help if you show your appreciation when you do get a reply (like this topic). If people answer your questions but you don't get back to them, they'll be less willing to answer further questions ;-)

Cheers,
Gijs
Airports: EHAM, EHLE, KSFO
Aircraft: 747-400
User avatar
Gijs
Moderator
 
Posts: 9544
Joined: Tue Jul 03, 2007 3:55 pm
Location: Delft, the Netherlands
Callsign: PH-GYS
Version: Git
OS: Windows 10

Re: Flight Dynamics Model

Postby Hooray » Sat Jul 07, 2012 4:12 pm

Please don't send support requests by PM, instead post your questions on the forum so that all users can contribute and benefit
Thanks & all the best,
Hooray
Help write next month's newsletter !
pui2canvas | MapStructure | Canvas Development | Programming resources
Hooray
 
Posts: 12707
Joined: Tue Mar 25, 2008 9:40 am
Pronouns: THOU

Re: Flight Dynamics Model

Postby awtudor » Mon Jul 09, 2012 9:39 pm

Hi,

Thanks everyone for the help. I have another question you might be able to answer. I am trying to generate a block diagram that looks something like this: http://www.mathworks.com/help/toolbox/a ... mple14.png
But I can not find the following two blocks (I have all the others)

"Unpack net_ctrl Packet from FlightGear "
http://www.mathworks.com/help/toolbox/a ... tgear.html

"Receive net_ctrl Packet from FlightGear"
http://www.mathworks.com/help/toolbox/a ... tgear.html

Should they all be in the same place? Am I forgetting to install something?

Gijs,

Ideally I would like some sort or realistic aircraft. It would be nice to also know a little about how the FDM was generated. However, this is not the main point of the project so something reasonable should work. As for the UAV, I only mention that because there is a lot of research in that area. If no models exist right now, that is fine.

Thanks,
Andy
awtudor
 
Posts: 5
Joined: Mon Jun 25, 2012 7:48 pm

Re: Flight Dynamics Model

Postby awtudor » Mon Jul 09, 2012 11:35 pm

Hi,

I think I figured it out. According to here: http://www.mathworks.com/help/toolbox/a ... tgear.html
"The Unpack net_ctrl Packet from FlightGear block unpacks net_ctrl variable packets received from FlightGear and makes them available for the Simulink environment.

Supported FlightGear versions:

v0.9.3

v0.9.8/0.9.8a

v0.9.9

v0.9.10

v1.0

v1.9.1

v2.0"

Currently I am running 2.6.0 so I think that could be the reason the blocks are not appearing. I will update after 2.0 installs.

Thanks,
Andy
awtudor
 
Posts: 5
Joined: Mon Jun 25, 2012 7:48 pm

Re: Flight Dynamics Model

Postby hvengel » Tue Oct 30, 2012 9:39 pm

Gijs wrote in Sat Jul 07, 2012 11:20 am:Hi Andrew,

sounds like an interesting project! I've tried to answer your questions:

1) The detailed models usually have some references inside the FDM files (in Aircraft/p51d/p51d-jsbsim.xml for example, you'll find some comments like "from NACA P-51X and P-51B test data" or "The rest of the drag functions are guesses."). JSBSim supports some "reference/documentation tags", but it isn't used very often, in the FlightGear aircraft at least. But the comments as the P51D has them work basically the same. The P51D has one of the better FDMs btw.

...

Cheers,
Gijs


Since the OP asked about how the models are built and where the data comes from I can provide some details with respect to the P-51D JSBSim FDM model. As the comments quoted above indicate some of the data comes from actual aircraft test data from NACA (mostly) and from other official sources such as the USAAF and the Royal Air Force. Also included in this is data gleamed from pilots manuals and training materials from that era (fuel consumption is one example).

The NACA data is contained in perhaps 8 to 10 reports that tested a wide range things like landing characteristics, stall characteristics, rudder control characteristics, high speed dives, rate of climb, speeds at various altitudes and power settings and so on. Also included in this are actual wing section lift/drag wind tunnel test data.

The P-51 series is somewhat unusual in that at the time it was by far the most tested aircraft in existence and various organizations such as NACA generated reams of data for the aircraft which were compiled into various test reports. Much of this is freely available on the NASA data server and is in the public domain. The test data is contained in about 200 pages of reports.

Even with this extensive data set there is much about the aircraft that is not well documented. I will give two examples:

1. The P-51D used NACA derived exhaust nozzles that generated thrust much the same way a rockets does (IE. the high velocity exhaust gases act like a rocket/jet). There is some test data on these nozzles available from NACA but tests used a Spitfire V (earlier less powerful model of the Merlin engine) or a P-40 (Allison engine) and a lot of reading between the lines was needed to model this on the P-51D.

2. The P-51 series is somewhat unique in that it was perhaps the only production aircraft to successfully apply the "Meredith Effect" which uses the heat from the cooling system to generate jet thrust. But there is no test data for this available, because wind tunnels of the needed size and velocity did not exist, other than some data points published in an artificial written by the chief designer, Lee Atwood see http://aviationclub.aero/article/view/3756/, in the late 1980s where he credits the "Meredith Effect" with much of the aircraft performance and a late 1930s NACA paper of a wind tunnel test using an electrically heated test device to simulate the "Meredith Effect" which was more or less a proof of concept (IE. was it possible - answer YES).

At best the modeling of these two things is more or less an educated guess that gives us an approximation of how these affected the aircraft thrust/drag. This is also true for some of the other "data" used in the FDM. In reality how well a particular aircraft is modeled is a function of two things.

1. The quality of the data available. In some cases like the P-51 there is extensive data available in other cases very little actual test data can be found.

2. How hard the person doing the modeling works on making the FDM as correct as possible. Not much more to add to this other than that this depends on #1 above and if the quality/quantity of data is lacking there is only so much that can be done to create an accurate FDM model.
hvengel
Retired
 
Posts: 1127
Joined: Sun Dec 24, 2006 5:35 am
Location: Minden Nevada


Return to The FlightGear project

Who is online

Users browsing this forum: No registered users and 6 guests